{"id":"0f045e92-8ede-4f4c-b026-df9353438058","arxiv_id":"1908.05430","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Baikal-GVD reports three candidate cascade events above 100 TeV and a diffuse neutrino flux upper limit about three times the IceCube flux.","lead":"Baikal-GVD, a neutrino telescope under construction in Lake Baikal, found three high-energy cascade events in its 2016 and 2018 data. The new upper limit on the astrophysical neutrino flux is about three times above the flux measured by IceCube, showing the young detector already approaches the sensitivity needed to see astrophysical neutrinos.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The three events above 100 TeV are not shown to be astrophysical: the paper itself says the atmospheric muon/neutrino background probabilities 'are in progress,' so the candidate claim and the 'same ballpark' conclusion rest on an unquantified background.","rationale":"The reader's weakest_assumption identifies exactly the load-bearing gap: the three events above 100 TeV are interpreted as astrophysical candidates without a computed atmospheric muon background. The paper's own sentence, 'The calculations of the probabilities to obtain such high multiplicity events from atmospheric muons and neutrinos are in progress,' is an explicit limitation that the stress-test rule requires flagging. This is not a manufactured concern; it is the difference between a calibrated search and an event count. The reader's verdict of CONDITIONAL is appropriate: the paper is an honest preliminary report with useful detector information and MC-based expectations, but the missing background estimate prevents a mature assessment. I considered whether the unsubtracted upper limit alone could justify the 'same ballpark' claim even if all three events are background, because an upper limit computed without background subtraction is conservative. However, the paper explicitly calls the three events 'promising high-energy cascade events - candidates for events from astrophysical neutrinos,' so the concern affects the central claim, not merely the limit's numerical value. The concrete test—estimating the atmospheric background through the full selection chain or a data sideband—would settle whether this concern lands. Verdict remains CONDITIONAL, pending that background estimate and a recomputed limit with proper background treatment.","tokens_in":5303,"tokens_out":4602,"duration_ms":48533,"concrete_test":"Run the full Baikal-GVD atmospheric muon and neutrino Monte Carlo through the identical cascade reconstruction and selection chain (Nhit > 20, E > 100 TeV, containment and quality cuts) for the 872 cluster-days, and cross-check with a data-derived sideband such as events reconstructed outside the instrumented volume or from randomized trigger times. If the predicted atmospheric background above 100 TeV is below about 0.2 events, the candidate interpretation stands; if it is comparable to the three observed events, the upper limit must be recomputed with background subtraction and the 'same ballpark' conclusion should be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that Baikal-GVD, at partial construction, already sees a neutrino flux 'only a three higher than the IceCube flux' and is 'in the same ballpark.' Section 3 bases this on three events with Nhit > 20 and reconstructed energy above 100 TeV in 872 cluster-days. The load-bearing step is the implicit identification of these events as astrophysical neutrino candidates. The paper explicitly concedes the background estimate is missing: 'The calculations of the probabilities to obtain such high multiplicity events from atmospheric muons and neutrinos are in progress.' Only the atmospheric-neutrino component is quantified (0.08 events/yr above 100 TeV, so about 0.19 events in this exposure); atmospheric muon bundles, which dominate the sample below 100 TeV, are not. All three candidates sit near the energy threshold (107, 153, 155 TeV), and only one is described as contained, so a misreconstructed or partially contained muon bundle could populate the final selection. Without a quantitative atmospheric background estimate, the events cannot be called astrophysical candidates, and the conclusion that GVD is already probing the IceCube flux is unsupported. The upper limit itself is formally conservative if the observed events are treated as signal with no background subtraction, but the paper's interpretive claim—that these are first astrophysical neutrino candidates—depends on the background being negligible.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a search for high-energy cascade neutrino events with the partially constructed Baikal-GVD detector, using 872 cluster-days of live time from one cluster in 2016 and three clusters in 2018. After a standard two-step reconstruction (vertex fit, then maximum-likelihood energy/direction fit) and quality cuts, the analysis selects 417 events with hit multiplicity Nhit > 13, of which 18 have Nhit > 20 and 3 have reconstructed energy above 100 TeV. The paper derives a 90% CL upper limit on a single-flavor E^-2.46 diffuse astrophysical neutrino flux using the Feldman-Cousins prescription [9], taking the three events as signal, and reports that the limit is only about a factor of three above the IceCube flux. The central interpretive claim is that Baikal-GVD, at partial construction, is already probing the astrophysical neutrino flux and that these three events are candidates for astrophysical neutrinos.","tokens_in":5592,"tokens_out":3691,"duration_ms":39812,"significance":"If the three events are indeed astrophysical neutrino cascades, this is an important early result: it would show that a modular, under-construction km^3-scale detector can approach IceCube sensitivity before completion. The analysis uses standard and reproducible tools: Monte Carlo simulation of signal and background, a well-defined reconstruction chain, and a conventional upper-limit prescription with an externally fixed spectral index. The paper is honest about its preliminary nature and explicitly states that the critical background probability calculation is still in progress. However, because that background calculation is the load-bearing element for the candidate interpretation, the significance of the result currently rests on an unquantified assumption rather than on demonstrated background rejection. The paper's value is therefore as an early sensitivity demonstration and a progress report, with the astrophysical interpretation not yet established.","major_comments":[{"comment":"The paper explicitly states that 'The calculations of the probabilities to obtain such high multiplicity events from atmospheric muons and neutrinos are in progress,' and it quantifies only the atmospheric-neutrino background (0.08 events/yr above 100 TeV), not atmospheric muon bundles, which dominate the event population below 100 TeV according to Fig. 2. Only one of the three selected events is described as contained, and the reconstructed energies are 107, 153, and 155 TeV, i.e., all three are just above the 100 TeV threshold. Without a quantitative estimate of the atmospheric muon-bundle background in this final sample, the identification of these three events as astrophysical neutrino candidates is not established. This identification is the load-bearing step for the paper's claim that Baikal-GVD is 'in the same ballpark' as IceCube and that these are 'candidates for events from astrophysical neutrinos.' The upper limit is conservative if the events are treated as signal, but the candidate interpretation requires the background to be negligible, which is precisely the unquantified point.","section":"Section 3, Table 2, Conclusion"},{"comment":"The paper quotes an energy resolution of about 30% for cascades but does not propagate this resolution into the selection or the limit. All three candidates are within a factor of 1.5 of the 100 TeV threshold (107, 153, and 155 TeV), so an atmospheric event with true energy below 100 TeV could migrate above the cut, while a genuine astrophysical neutrino at the threshold could migrate below. This migration matters because the expected atmospheric background rises steeply with decreasing energy, as shown in Fig. 2, and because the upper limit is based on the number of observed events above threshold. Please provide an estimate of the migration effect, either from unfolded MC distributions or from a smeared signal-plus-background model, or justify quantitatively that the effect is negligible for the reported limit.","section":"Section 2.2, Table 2, Section 3"},{"comment":"The 90% CL upper limit is derived from three observed events with no background subtraction and no systematic uncertainties assigned to the effective area, the energy scale, or the Monte Carlo simulation. The paper should state explicitly that the limit is a conservative upper limit under the assumption that all three events are signal, and that it is not an astrophysical flux constraint if a substantial fraction of the events is atmospheric. This clarification is necessary because the text moves from the limit to the claim that the detector is 'in the same ballpark' as IceCube without distinguishing a flux measurement from an upper limit with unquantified background.","section":"Section 3, Fig. 7 (right)"}],"minor_comments":[{"comment":"This sentence should read 'a factor of three higher than the IceCube flux'; the current phrasing is unclear and should be corrected.","section":"Section 3, 'Our limit is only a three higher...'"},{"comment":"References [4] and [5] are identical (Astropart.Phys. 25, 140 (2006)); one of them is likely intended to be a different paper and should be corrected.","section":"References"},{"comment":"The introduction says that four additional clusters commissioned in 2017-2019 bring the total to 1440 optical modules, while the conclusion says that five clusters with 1140 OMs are taking data since April 2019. These numbers are inconsistent and should be reconciled.","section":"Intro and Conclusion"},{"comment":"The paper states that one of the three candidates is a contained event but does not define the containment criterion. Please specify how containment is determined (e.g., distance from cluster boundary or vertex position relative to instrumented volume).","section":"Section 2.2, 'contained event'"},{"comment":"The right panels show gamma-ray source sky maps with 2-degree circles around the reconstructed event directions, but the figures lack clear axis labels and a legend for the color scale. The reader cannot tell whether the circles are centered on the sources or on the reconstructed directions, or what correlation is being claimed.","section":"Figures 5 and 6"},{"comment":"Equation (1) has a nonstandard layout with the normalization written as 1/(N_hit - 4) inside the sum; while the meaning is understandable, the notation should be cleaned up for publication.","section":"Section 2.2, Eq. (1)"}],"recommendation":"major_revision","confidential_remarks":"This is an ICRC proceedings-style paper, and the community understands its preliminary status. The central weakness is openly acknowledged by the authors: the atmospheric background probability calculation is 'in progress.' The result can become publishable either by adding the quantitative background estimate (including muon bundles) or by reframing the conclusion as an observation of three candidate events with a conservative upper limit and clearly stating that the astrophysical interpretation is not yet established. The energy-resolution migration issue also requires a quantitative response. I recommend major revision rather than rejection because the selection procedure and limit computation are standard and the missing piece is identifiable and fixable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"What you should know: this is a real measurement, not a simulation. Using 872 cluster-days of Baikal-GVD data from 2016 and 2018, the collaboration finds three events with reconstructed energy above 100 TeV and Nhit > 20, and derives a 90% CL single-flavor E^-2.46 upper limit about three times the IceCube flux. The paper explicitly says the atmospheric muon/neutrino background probabilities for these three events are still being calculated. That is the key caveat, and the authors deserve credit for putting it in the text rather than burying it.\n\nWhat is good: the data are new, the three candidate events are new, and this is the first cascade-channel search from a partially built GVD. The analysis pipeline is standard — vertex reconstruction by time minimization, energy/direction by maximum likelihood, then quality cuts — but it is applied to a new detector and gives an independent cross-check in the Northern Hemisphere. The spectral index is fixed to the IceCube value, no free parameter is fitted to force agreement, and the limit is a straightforward Feldman-Cousins count. There is no circularity here. The authors also openly label the result as preliminary, which is the right tone for an ICRC proceedings.\n\nWhere it is soft: the load-bearing gap is the missing atmospheric muon bundle background. The atmospheric neutrino background is small (about 0.08 events/yr above 100 TeV), but atmospheric muons dominate below 100 TeV and are not quantified at all. All three candidates sit close to the energy threshold (107, 153, 155 TeV), and only one is described as fully contained. A misreconstructed or partially contained muon bundle could easily populate that region. The quoted 30% energy resolution is also not propagated into the limit, though for an upper limit based on counts this likely would not change the conclusion dramatically. Minor points: the optical module count is inconsistent (1440 in the introduction vs 1140 in the conclusion), and references [4] and [5] are duplicated. None of these are fatal, but the background estimate is essential before the events can be called astrophysical candidates.\n\nWho should read it: people working on neutrino telescopes, especially those tracking Baikal-GVD progress and early science. It is a status report, not a discovery claim, and it should be read as such. I would send it to peer review: the experimental result is worth refereeing, with the clear instruction that the background probabilities need to be added or the candidate interpretation softened. The paper is honest, the work is reproducible in principle, and the gap is stated by the authors themselves — that is exactly the kind of paper a good referee can help improve.","headline":"A genuinely new but still preliminary GVD cascade search: three candidates above 100 TeV and a limit within a factor of three of IceCube, with the paper itself admitting the atmospheric background estimate is not done yet.","tokens_in":6427,"tokens_out":1602,"would_cite":true,"duration_ms":18239,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A partially built Baikal-GVD already sets a neutrino flux upper limit only about three times above IceCube's.","keywords":["neutrino astronomy","Baikal-GVD","cascade events","astrophysical neutrinos","diffuse flux","upper limit","Cherenkov detector","high-energy neutrinos"],"falsifier":"Compute, via Monte Carlo simulation, the expected number of atmospheric muon and neutrino background events in 872 cluster-days passing the same $N_{\\mathrm{hit}}>20$ and $E>100~\\mathrm{TeV}$ selection; if that expectation is of order one or more, the three candidates are consistent with atmospheric background and the astrophysical interpretation fails.","tokens_in":5158,"feed_emoji":"🔭","tokens_out":6492,"duration_ms":61136,"temperature":0.7,"pith_summary":"The paper's goal is to show that Baikal-GVD, a kilometer-scale neutrino telescope still under construction in Lake Baikal, can already search for astrophysical neutrinos through cascade events. Using 872 cluster-days of data recorded by one cluster in 2016 and three clusters in 2018, the analysis selected 417 cascade-like events, of which 18 had more than 20 hit optical modules and three had reconstructed energies above 100 TeV. Taking those three events as neutrino candidates, the collaboration derives a 90% confidence single-flavor upper limit on an $E^{-2.46}$ diffuse flux that is about three times higher than the flux measured by IceCube. The authors conclude that the detector is already in the same sensitivity ballpark as IceCube and, with more data and refined analysis, should identify the first astrophysical neutrinos. The result is explicitly preliminary: the probability that atmospheric muons or neutrinos produce the three events is still being calculated.","feed_headline":"Baikal-GVD finds 3 neutrino candidates, nears IceCube flux","feed_subtitle":"Using 872 cluster-days, the detector's first cascade search reaches within a factor of three of IceCube.","key_machinery":"The load-bearing mechanism is the cascade-event selection and two-step reconstruction chain. A cascade event is a neutrino interaction that produces a shower of particles and a roughly spherical Cherenkov light front, and the analysis isolates such events by demanding a high hit multiplicity, $N_{\\mathrm{hit}}>20$, and reconstructed energy above 100 TeV, where atmospheric backgrounds are strongly suppressed. In the first reconstruction step the shower vertex is fit by a chi-squared minimum using photon arrival times under the assumption that the shower is a point-like light source; in the second step the direction and energy are fit by maximum likelihood using amplitude probabilities calibrated against Monte Carlo simulations of Cherenkov light propagation in water. This machinery matters because it lets a sparse, partially deployed array retain energy resolution of about 30%, a median direction resolution of about 4 degrees, and a vertex resolution of about 2 meters, making early-construction data astrophysically competitive.","core_discovery":"The central claim is that cascade-mode neutrino searches with a partially built Baikal-GVD array reach the astrophysical flux scale measured by IceCube. From 872 cluster-days collected in 2016 and 2018, the collaboration reconstructed 417 cascade-like events, restricted to 18 with $N_{\\mathrm{hit}}>20$, and found three events with reconstructed energy above 100 TeV. Under the assumption of an $E^{-2.46}$ spectrum and single-flavor normalization, the three events are used to set a 90% confidence level upper limit on the diffuse astrophysical neutrino flux; the limit is only a factor of about three above the IceCube flux. The paper presents the three high-energy events as candidates for astrophysical neutrinos and states that more data and a refined analysis will allow the first astrophysical neutrinos to be identified.","pith_inferences":["If the three candidates are genuine astrophysical cascades, their zenith distribution offers an early glimpse at the Northern-sky neutrino sky, complementing IceCube's largely Southern-sky sensitivity.","The sensitivity could be improved ahead of full construction by stronger rejection of atmospheric backgrounds, for example using the outer strings as a veto, rather than by waiting for more deployed volume.","A dedicated search combining 2019's five-cluster data with the 2016 and 2018 samples could test whether the three events cluster in a particular sky region or are isotropically distributed."],"forward_implications":["If the limit is correct, a partially built Baikal-GVD can independently probe the same diffuse astrophysical neutrino flux that IceCube discovered, from the Northern hemisphere.","With the planned first-stage array of nine clusters and more live time, the three candidates should grow into a statistically significant excess, allowing the first claimed astrophysical neutrinos at Baikal-GVD.","The 90% CL limit, being only a factor of three above the IceCube flux, provides a useful cross-check on the IceCube measurement in an energy range around 100 TeV.","More data will allow the atmospheric background probability to be quantified, turning the current candidates into either a confirmed signal or a background-dominated sample."],"supporting_citations":[{"why":"Supplies the detector layout and clustering concept that define the simulated response.","marker":"[1]"},{"why":"Provides the IceCube discovery of a diffuse astrophysical neutrino flux that the search aims to confirm.","marker":"[2]"},{"why":"Fixes the assumed E^-2.46 spectrum and single-flavor normalization used to compute expected signal and the upper limit.","marker":"[6]"},{"why":"Establishes the cascade reconstruction and event-selection criteria, including the Nhit > 20 cut and expected resolutions, used to classify candidates.","marker":"[7]"},{"why":"Supplies the IceCube diffuse-flux measurement used as the comparison normalization in the expected-event distributions.","marker":"[8]"},{"why":"Gives the statistical prescription used to convert the three observed events into a 90% CL upper limit.","marker":"[9]"}],"fun_headline_variants":["Baikal-GVD sees 3 neutrino cascades, within 3x of IceCube","3 cascade events in Baikal-GVD reach IceCube flux scale","Baikal-GVD's first cascade search nears IceCube flux","Three neutrino candidates in Baikal-GVD, flux near IceCube"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result collapses if the three recorded high-energy events are actually atmospheric muons or atmospheric neutrinos rather than astrophysical neutrinos; the paper explicitly notes that the calculation of the probabilities for obtaining such high-multiplicity events from atmospheric backgrounds is still in progress.","fun_headline_variants_meta":{"raw":{"variants":["Baikal-GVD sees 3 neutrino cascades, within 3x of IceCube","3 cascade events in Baikal-GVD reach IceCube flux scale","Baikal-GVD's first cascade search nears IceCube flux","Three neutrino candidates in Baikal-GVD, flux near IceCube"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000746,"raw_usage":{"total_tokens":3255,"prompt_tokens":807,"completion_tokens":2448,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":423,"completion_tokens_details":{"reasoning_tokens":2368}},"tokens_in":423,"tokens_out":2448,"duration_ms":15472,"temperature":1.0,"reasoning_tokens":2368,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:13:53.721470+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute, via Monte Carlo simulation, the expected number of atmospheric muon and neutrino background events in 872 cluster-days passing the same $N_{\\mathrm{hit}}>20$ and $E>100~\\mathrm{TeV}$ selection; if that expectation is of order one or more, the three candidates are consistent with atmospheric background and the astrophysical interpretation fails.","supporting_citations":[{"cited_title":"Aynutdinov et al., NIM A742 82-88 (2014)","cited_arxiv_id":null,"evidence_quote":"Supplies the detector layout and clustering concept that define the simulated response."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Fixes the assumed E^-2.46 spectrum and single-flavor normalization used to compute expected signal and the upper limit."},{"cited_title":"Avrorin et al., PoS (ICRC2017)962, (2017)","cited_arxiv_id":null,"evidence_quote":"Establishes the cascade reconstruction and event-selection criteria, including the Nhit > 20 cut and expected resolutions, used to classify candidates."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the IceCube diffuse-flux measurement used as the comparison normalization in the expected-event distributions."},{"cited_title":"Feldman and R.D","cited_arxiv_id":null,"evidence_quote":"Gives the statistical prescription used to convert the three observed events into a 90% CL upper limit."}],"review_version":1}